Composites of an organic material and an inorganic material have been under various studies. Compounding an inorganic filler into an organic polymer, modifying a metal surface by coating with an organic polymer, and like techniques have been made use of on an industrial scale. The organic and inorganic materials composing such composites can admittedly bring about unexpected improvement on certain physical properties. Nevertheless, because the individual materials composing an organic-inorganic composite are of sizes of the order of micrometer or greater, improvement on many other physical properties and performance properties that can be obtained by use of the composite is nothing more than the levels predictable from the performance and physical properties of the constituent materials based on sum rule.
Recently, there has been an increase of research into organic-inorganic composites in which organic and inorganic materials are combined with their domain sizes of the order of nanometer or even on molecular level. Such composites are promising for exhibiting not only the characteristics and merits inherent to the individual materials but also new functionality that is quite different from the functions of the individual materials and therefore unpredictable by sum rule.
Such organic-inorganic composites include chemically bonded composites in which one of the constituent materials is bonded to the other on molecular level via a covalent bond and dispersed composites in which one of the materials is finely dispersed in the other (matrix). A sol gel process is widely utilized to synthesize inorganic materials for use in these organic-inorganic composites. A sol gel process is a technique for obtaining a crosslinked inorganic oxide in low temperature, in which precursor molecules are hydrolyzed followed by polycondensation. The problem of the inorganic materials prepared by the sol gel process is poor storage stability such that gelation occurs in a short time. Non-Patent Document 1 reports an attempt to improve storage stability with attention focused on dependence of the condensation reaction rate of an alkyltrialkoxysilane on the alkyl chain length. According to the report, polycondensation of methyltrimethoxysilane is followed by adding a long-chain alkyltrialkoxysilane having a low condensation reaction rate thereby to block the silanol groups of the polysiloxane. The report also teaches that methyltrimethoxysilane is polycondensed using an aluminum catalyst and, when a predetermined molecular weight is reached, acetylacetone is added to the reaction system whereby to cause a ligand exchange reaction in the system. However, these methods are still insufficient for improving the storage stability. Furthermore, inorganic materials synthesized by a sol-gel process have a disadvantage of poor flexibility.
Patent Document 1 discloses a curing composition containing a specific silicon containing polymer as a chemically bonded organic-inorganic composite. However, the curing composition containing the silicon containing polymer has insufficient performance. In particular, a cured product of the curing composition is unsatisfactory in heat resistance and handling properties.    Patent Document 1: JP-A-2002-356617    Non-Patent Document 1: The Chemical Society of Japan, Nippon Kagaku Kaishi (Bulletin of the Chemical Society of Japan), No. 9, 571 (1998)